Mobile Edge Computing-based Task Cooperative Scheduling And Resource Allocation Strategy

The rapid development of mobile networks and smart devices promotes the emergence of a large number of resource-intensive applications,such as cloud gaming,virtual reality(AR)and image recognition.These applications not only consume a lot of memory and computing resources,but also have high requirements for delay.However,due to the limitations of battery capacity and computing power,it is difficult for mobile devices(MDs)to meet the delay requirements of these tasks.To tackle this issue,mobile edge computing(MEC)technology is recently proposed.By deploying high performance MEC servers at the edge of radio access networks(RAN),e.g.,the base station(BS)of cellular networks,and conducting computation task execution for the MDs,the efficiency of task execution can be effectively improved,and the strict delay requirements of user tasks can be satisfied.This thesis mainly studies the task offloading,task scheduling,and resource allocation problems of MEC systems,including the following specific contents:Firstly,the MEC technology is briefly introduced,and the current research works on task offloading,task scheduling and resource allocation algorithms in MEC system at home and abroad are summarized and analyzed in detail.For an MEC system consisting of a number of MEC servers and one MD which generates a series of tasks characterized by their dependency relationships,considering the fairness among tasks,the system delay is defined as the maximum completion time of all tasks in the system,and considering constraints including task computing mode selection,slot allocation and MEC servers computing resources,the joint task offloading and computation scheduling problem is formulated as a worst-case latency optimization problem which minimizes the maximum completion time of all the tasks.As the original optimization problem is a integer nonlinear programming(INLP)problem,which can not be solved conveniently.We propose a task priority-based dynamical offloading and computation scheduling algorithm,and established a virtual time axis,for dynamically changing lower priority tasks,a task weight and data size-based task offloading and computation scheduling algorithm is proposed,and a multiple knapsack-based heuristic algorithm is proposed for high priority tasks.Then,the strategy of joint task offloading and scheduling is obtained.Jointly considering the characteristics of the computing resources of MEC servers,the tasks of MDs and user mobility,a utility optimization-based joint task offloading and resource allocation algorithm is proposed for blockchain-enabled edge-cloud collaboration systems.By defining the utility function as the weighted sum of task execution utility and blockchain utility in the system,and considering the constraints including computing mode selection,the transmission rate and the maximum delay tolerance of tasks,etc,the joint task offloading and resource allocation problem is formulated as the system utility maximum problem.In order to solve the problem,a heuristic algorithm is proposed.Firstly,Kuhn munkres(K-M)algorithm is applied and a migration selection algorithm is proposed to determine the joint task offloading and migration selection strategy.Then,the Lagrange dual method and gradient descent iterative algorithm are used to determine the resource allocation strategy.Finally,an iterative method is carried out to determine the joint task computing mode selection,migration selection,block generation selection and resource allocation strategy.Simulation results show that the proposed algorithm offers better system performance than previous schemes.